Published Paper Details:

The goal of this project is to create an automated system for detecting cyberbullying on social media platforms using Natural Language Processing (NLP) for text classification and Machine Learning for behavioral analysis. Cyberbullying, a growing issue in online interactions, often leads to significant psychological and emotional harm, especially among young users. This project employs NLP techniques to identify and classify harmful content, distinguishing it from neutral or positive interactions based on linguistic patterns, sentiment, and tone. Text data from various social media sources, annotated with labels for bullying and non-bullying content, serves as the dataset. The first phase involves data preprocessing, including text cleaning, tokenization, and word embedding techniques such as Word2Vec or BERT embeddings to capture semantic meaning. With these prepared data inputs, machine learning models such as Support Vector Machines (SVM), Random Forest, or deep learning models like Recurrent Neural Networks (RNNs) and Transformers are trained to classify text accurately. To enhance detection accuracy, this project also includes behavioral analysis by examining user interaction patterns, frequency of harmful words, and network analysis to track the dynamics of cyberbullying incidents. Integrating these analyses enables the model to detect complex bullying patterns that rely not only on language but also on behavioral indicators. This system aims to provide social media platforms with a reliable tool to flag and mitigate cyberbullying, promoting safer online environments.